Explain consciousness to the layman.

[Mr. Scott]

The fact is that you THINK current computers with current technology of software and hardware are capable of achieving consciousness.

I do not.

Oh look, you altered my statement to make your point! Why did you feel the need to do that?

I never said "current computers with current technology of software and hardware." I said I thought it in theory possible, though astronomically difficult, to build a conscious machine.

Nevertheless, I want to read your succinct and clear explanation of why you do NOT think current computers with current technology of software and hardware are capable of achieving consciousness. Is the limit qualitative or quantitative?

I await your response.

 

[rocketdodger]

This is fairly interesting.

Along with something at least close to "measurable" it provides decent working definitions for things like emotions.

http://www.consscale.com/en/levels/description.html

 

[Belz...]

A computer could no more be programmed to have thoughts, memories, feelings or sensations than the printing press could be set up to print them. No such programs exist, and it is pure conjecture to ever suppose that they could exist.

Seems to me like we're assuming that thoughts, memories, feelings and sensations are somehow "special".

Oh, well, if you've taught courses.

Ah, the anti-elitist "knowledge is bad" line of argument ?

 

[Belz...]

If you really believe that, there's not much I can do to change your mind.

I will just have to study harder. Clearly if I did a bit more work on the subject I'd learn all about those programming techniques for producing feelings. I suppose Mr Scott won't be telling me about it, but I can just listen in for when he tells somebody else.

You say that Scott's conclusion is a belief, and then you just argue from "is just so". Fascinating.

 

[punshhh]

It *may* not, but if you wanted a confidence value I would put it at like 99%.

Degrees of certainty are irrelevant, it remains a possibility we cannot eliminate.

That is, I am 99% certain that a kind of matter which merely obeys the mathematically consistent modeling of physicists is enough for life to be possible.

Again the percentage of certainty is irrelevant, we can only assume that our modeling of matter is comprehensive enough to for example simulate a physical system such that consciousness will emerge. When we get close to emulating consciousness we will be able to fumble our way through by trial and error.

Life is different because of how it behaves according to those mathematically consistent rules, not because it needs some special rules in the first place.

Did DNA develop spontaneously in nature in a rock pool?
If so why have other groupings of atoms as complex and codified as DNA not also arisen naturally and resulted in complex groupings analogous to living things all around us?

 

[Belz...]

"What computers are built from" is not the issue.

What the brain is built from is the issue, since brains are conscious and computers are not.

Wait, wait. You just assert that they are not, and then from this conclude that therefore what computers are made of is not important. How is what brains made of important, then, since we already know that computers are not conscious ?

Oh, wait. Maybe consciousness has nothing to do with what things are made of but how they behave, just like calculators.

 

[Belz...]

(except when it isn’t)

When isn't life chemical ?

 

[Belz...]

Sadly, no. There really are programmers out there - professional programmers, inasmuch as their job title is "programmer" and they actually get paid - who have not the faintest notion how computers work, nor that instructions are causally related.

I've worked with such people. Well, more accurately, I've carefully avoided such people until they took the hint and went into real estate or advertising.

I don't doubt for a moment that Westprog is a programmer.

Well, I'm a programmer and I have no clue about electronics. But I know how computer programs operate. Westprog doesn't.

 

[Belz...]

In answer to the above I post this reply

"it's important to be able to support your arguments with more than assertions so...

...how do you know ?"

Would you mind making some sense ?

Thank you.

 

[rocketdodger]

Did DNA develop spontaneously in nature in a rock pool?
If so why have other groupings of atoms as complex and codified as DNA not also arisen naturally and resulted in complex groupings analogous to living things all around us?

I would argue that there are no other such systems that could have arisen in the Terran environment with even close to the same likelihood.

 

[westprog]

How does that answer my post ? You think there is no causal relation between instructions, then you are wrong, plain and simple, or you are using an unrealistic definition of "relation". If you had ever programmed at all, you'd realise that.

A causal relationship between the state of a physical system and it's predecessor is inherent in all such systems. That's not some special trick of computers, it's common to everything in the universe.

The whole area of parallelising code relies on the fact that succesive instructions might have nothing to do with each other.

 

[westprog]

Protip: it's important to be able to support your arguments with more than assertions so...

That would pretty much kill off this thread altogether if applied equably. Don't suppose it will be though.

...how do you know ?

I would express it as - we have no more reason to suppose that computers are conscious than any other inanimate object.

 

[westprog]

Tonight's Horizon programme may be of interest:

Marcus du Sautoy, the Simonyi Professor for the Public Understanding of Science at the University of Oxford, presents Horizon: The Hunt for AI on BBC Two at 21:00 BST on Tuesday 3 April. Watch online afterwards (UK only).

It's possible to use software like Tunnelbear to convince the BBC that you live in England. I don't condone this practice but it certainly happens.

I suspect they'll conclude that human-like intelligence/consciousness is too difficult, and is unnecessary; that we should be considering different, complementary forms of artificial intelligence to assist us in areas we find problematic.

 

[westprog]

Seems to me like we're assuming that thoughts, memories, feelings and sensations are somehow "special".

The desperate need for nothing about human beings to be "special" doesn't mean that it's possible to claim the ability to program thoughts, memories, feelings and sensations, when nobody knows how to do that.

Ah, the anti-elitist "knowledge is bad" line of argument ?

No, it's the "I'm an expert on this subject therefore I am right and you are wrong argument".

I'd prefer to see expertise demonstrated rather than asserted. I'm regularly amused to see how much effort some people put into trying to debunk other people's backgrounds and assert their own. Amazing how many people posting here seem to be simultaneously trained biologists, neurologists, AI programmers and philosophers. It means they can just post

no

and expect to be believed.

 

[dlorde]

This is fairly interesting.

Along with something at least close to "measurable" it provides decent working definitions for things like emotions.

http://www.consscale.com/en/levels/description.html

Yes; I'm not sure I agree with every detail, but it's a positive and constructive approach.

 

[westprog]

Well, I'm a programmer and I have no clue about electronics. But I know how computer programs operate. Westprog doesn't.

Yes, there are a lot of programmers out there with varying degrees of expertise. Some have written sort routines in Javascript. Some have patched machine code in running systems. Not many have had to think very deeply about their assumptions about how their machines work.

Here's a simple test - if you can change the order of instructions without changing the outcome of a program, then you can legitimately assume that there is no causal relationship between said instructions.

 

[westprog]

You say that Scott's conclusion is a belief, and then you just argue from "is just so". Fascinating.

Mr Scott made a claim - that feelings could be programmed in a computer. I don't find it strange that this claim was allowed to pass unprotested by people who support the computationalist view, or that my objection to the statement should draw immediate criticism - including the assertion that I clearly know nothing about computers.

I will point out, however, that neither Mr Scott or anyone else, AFAIAA, has shown exactly how to program feelings into a computer. It takes a bit more than just saying that they aren't "special".

 

[dlorde]

Did DNA develop spontaneously in nature in a rock pool?

Very probably not.

If so why have other groupings of atoms as complex and codified as DNA not also arisen naturally and resulted in complex groupings analogous to living things all around us?

It's quite possible that the earliest replicators were RNA-based, and were succeeded by DNA-based organisms that incorporated their RNA chemistry.

Organic chemistry is really the only choice for an Earth-like environment; other environments might potentially favour different elements, but it's speculative. There may have been other organic replicators, but once a particularly efficient and successful form arises (and DNA/RNA organisms are particularly efficient and successful), they probably become food; or maybe they weren't sufficiently robust, or perhaps they didn't get started in time, or perhaps the right circumstances never arose.

 

[dlorde]

The whole area of parallelising code relies on the fact that succesive instructions might have nothing to do with each other.

Really? AIUI parallelising is running multiple independent processes concurrently - successive instructions of each process are directly related. If a single processor time-shares multiple processes, context-switching occurs explicitly to maintain the relationship. One could be pedantic, and include the process management and context switching instructions as part of that relationship (but that's really only useful from the OS or chip-microcode developer's perspective).

So can you explain more precisely what you mean by 'succesive instructions might have nothing to do with each other' ? An example would help.

 

[westprog]

Really? AIUI parallelising is running multiple independent processes concurrently - successive instructions of each process are directly related. If a single processor time-shares multiple processes, context-switching occurs explicitly to maintain the relationship. One could be pedantic, and include the process management and context switching instructions as part of that relationship (but that's really only useful from the OS or chip-microcode developer's perspective).

So can you explain more precisely what you mean by 'succesive instructions might have nothing to do with each other' ? An example would help.

x : = 1;
y := 2;
WRITELN(x+y);

or

y := 2;
x := 1;
WRITELN(x+y);

Clearly, the two fragments are equivalent in their outcome. The first and second instructions have no causal relationship. They could, if wished, be executed simultaneously on parallel processors. The result of the third instruction will be affected by the first two.

( I suppose that I should really do this in assembly code, but the principle is the same).